AI-100 : Preparing for the Azure AI Engineer certification
In a hurry ? Please skip to the part you want with the index below.
- Background
- Preparation
- Scheduling the exam
- Pre-exam process [Warning:Stress ahead!]
- The AI-100 exam
- Final thoughts
Background
“Knowledge is power. Knowledge shared is power multiplied.” - Robert Boyce
AI-100 is the Azure AI Engineer exam. I cleared the exam today ie Nov 1, 2020. This is my 3rd Azure certification. You can read about the others here.
I recently worked on Azure bot services for a project. It got me interested in Azure’s cognitive services. After my employer organized a related training session on AI-100, I got the drive to go for this certifrication.
Preparation
- Analyze solution requirements (25-30%)
- Design AI solutions (40-45%)
- Implement and monitor AI solutions (25-30%)
As you see - this needs a lot of foundation knowledge (Azure + data engineering). I found 50% of the exam questions to involve the fundamentals of Azure. The rest was a mix of cogntitive services and other supporting tools. The exam is targeted for a data engineer who is comfortable with Azure. Since I cleared AZ-300 and AZ-301 recently, I too found the exam to be easy . But it still involved going through a lot of course material !
There aren’t many AI-100 courses out there. But couple of websites which helped me prepare :
Cloud Academy
- I recommend the AI-100 learning path.
- Its well focused and short (5 hrs or so).
- The course focuses on Azure’s cognitive services and some of the supporting tools.
- Its expected for the audience to know the basics of Azure.
Udemy practice tests
- This offering helped me a lot.
- The 5 practice tests are really good and test your knowledge well. They also point out chinks in your Azure knowhow quite a bit.
- The duration of each test was around 1 hour each though I completed each test within 30 min.
Scheduling the exam
You can schedule the exam through the Microsoft Certification Dashboard page.
Pre-exam process [Warning:Stress ahead!]
Like AZ-301, this was a bit of a stressful experience for me. I did the pre-exam prep below around 20 min before the exam :
- I had to download a software which tested my machine for compatibility.
- Send pics from 4 directions of my surroundings (closed room).
- After a live 360 degrees scan from the Proctor, The exam was supposed to begin.
- The pre exam profile questionaire wrapped up quickly.
-
However here it got a bit stressful
- The software stopped working after it couldn’t stop some background app. It had me waiting for some response…but to no avail.
- The supervisor mentioned some tech issue on the Mac version of the software and gave me a voucher to give the exam some other time in the future.
- Scheduling the exam a 2nd time after 5 days didnt help. This time the app got stuck at ‘Loading’…and no amount of hitting the ‘retry’ button helped. Got another voucher again and rescheduled the exam after 3 days again.
The AI-100 exam
- It was an exam of 3 hours with 30 questions and passing score of 70%.
- Since I had completed the practice tests well in advance, I knew that time wasnt an issue for me.
- The practice test experience helped. Found a lot of similar questions from the tests.
- I submitted the exam in 30 min. Completed the questions in 25 min. Spent 5 min reviewing the questions about whom I wasn’t sure about.
- I had to fill a survey on myself and nature of the exam after this (no impact on the exam results).
- I immediately received the congratulatory message page for clearing the exam.
The exam report with certification id arrived in 10 minutes.
Final thoughts
- I loved studying for A1-100. It really improves your basics on data engineering and reference architectures.
- I think I might go for the data engineer path of certifications as my current work is pushing me towards it + I found it quite interesting when studying for AI-100.
- Besides the main cognitive services, I focused on smaller topics which helped :
- Azure reference architectures for AI (especially IoT and real time analytics use cases).
- Azure on the Edge.
- Machine learning concepts and tooling.
- Azure Stream analytics.
- HDInsight (Hadoop, Spark, Storm, Hive, Kafka etc).
- Data Factory and Data Bricks.
- Practice tests help a lot to identify chinks in your armour. Give as many as possible and multiple times if needed.
I wish you the best of luck if you plan on giving this exam .
Feel free to share your experiences. Every bit of knowledge helps .